This project develops a model of defensiveness as a precursor to unethical behaviors. Defensiveness is defined as the tendency to avoid or deny negative characteristics of the self, and is operationalized in terms of existing, well-established measures of self-deception and self-structure. The project applies an existing model of the evaluative organization of the self (i.e., compartmentalized versus integrative self-organization) to the novel domain of ethical behavior. We predict that a compartmentalized self-structure, in which individuals organize their positive and negative self-beliefs into distinct self-aspects, is often defensive, and allows individuals to avoid or deny negative self-knowledge. This defensive structure means that they also avoid the negative implications for the self of any unethical behaviors, so they may feel less inhibited. In contrast, persons with integratively-organized selves (positive and negative self-beliefs are closely associated within the same self-aspects) are less defensive - they acknowledge and confront negative self-attributes and therefore are likely to consider the negative implications of unethical behaviors for the self, which should deter them from those behaviors. The project consists of four empirical studies using college-student samples. The first two studies test the defensive qualities of compartmentalized versus integrative self-structures and examine their correlations with a variety of dishonest behaviors (including two laboratory cheating paradigms), taking into account existing models of self-deception and ego depletion. Two follow-up experiments use manipulations of threat or bolstering of the self to alter self-structure to be more or less defensive, with subsequent consequences for likelihood of dishonest behaviors. There are two long-term objectives: First, this project will introduce a conceptual model of defensiveness in terms of the (low) tendency to acknowledge and activate negative beliefs about the self, thereby operationalizing this clinically-based term for the purposes of empirical research. Second, this project introduces a broad-based model for ethical behavior, by demonstrating how the process of acknowledging negative self-attributes and beliefs may influence self-regulation. A better understanding of this process could lead to interventions that minimize defensiveness, for example by suggesting how people might alter their self-construals to acknowledge and confront negative attributes, thereby increasing the likelihood of honest behavior.